System and method for a cooperative conversational voice user interface

Abstract

A cooperative conversational voice user interface is provided. The cooperative conversational voice user interface may build upon short-term and long-term shared knowledge to generate one or more explicit and/or implicit hypotheses about an intent of a user utterance. The hypotheses may be ranked based on varying degrees of certainty, and an adaptive response may be generated for the user. Responses may be worded based on the degrees of certainty and to frame an appropriate domain for a subsequent utterance. In one implementation, misrecognitions may be tolerated, and conversational course may be corrected based on subsequent utterances and/or responses.

What is claimed is: 
     
       1. A computer-implemented method of facilitating natural language system responses utilizing a context stack generated based on multiple prior natural language utterances received, the method being implemented by a computer system that includes one or more physical processors executing one or more computer program instructions which, when executed, perform the method, the method comprising:
 tracking, by the computer system, a series of contexts respectively identified for a series of natural language utterances received by the computer system during a current conversation, the series of contexts comprising at least a first context identified for a first utterance at a first time and a second context identified for a second utterance at a second time after the first time; 
 generating, by the computer system, a context stack based on the tracked contexts comprising the series of contexts in reverse chronological order of a time of identification by the computer system, wherein the second context is listed before the first context; 
 receiving, at the computer system and as part of the current conversation, a third natural language utterance from an input device at a third time after the second time; 
 determining, by the computer system, whether the third natural language utterance corresponds to one or more of the series of contexts in the generated context stack by comparing the third natural language utterance to one or more of the series of contexts in the order the series of contexts are listed in the generated context stack; and 
 responsive to a determination that the third natural language utterance corresponds to one or more of the series of contexts in the generated context stack, interpreting, by the computer system, the third natural language utterance using the corresponding one or more contexts. 
 
     
     
       2. The method of  claim 1 , the method further comprising:
 generating, by the computer system, a response to the third natural language utterance based on the interpretation. 
 
     
     
       3. The method of  claim 1 , wherein determining whether the third natural language utterance corresponds to one or more of the series of contexts in the generated context stack comprises:
 identifying, by the computer system, one or more of the series of contexts in the generated context stack based on one or more recognized words of the third natural language utterance. 
 
     
     
       4. The method of  claim 1 , wherein determining whether the third natural language utterance corresponds to one or more of the series of contexts in the generated context stack comprises:
 determining, by the computer system, whether the third natural language utterance corresponds to a most recent context of the series of contexts in the generated context stack; and 
 responsive to a determination that the third natural language utterance does not correspond to the most recent context, determining, by the computer system, whether the third natural language utterance corresponds to a second most recent context of the series of contexts in the generated context stack. 
 
     
     
       5. The method of  claim 1 , the method further comprising:
 responsive to a determination that the third natural language utterance does not correspond to one or more of the series of contexts in the generated context stack, determining, by the computer system, a third context for the third natural language utterance based on one or more recognized words of the third natural language utterance; 
 updating, by the computer system, the generated context stack based on the third context by placing the third context first; 
 receiving, at the computer system, a fourth natural language utterance; 
 responsive to a determination that the fourth natural language utterance corresponds to one or more of the series of contexts in the updated context stack, interpreting, by the computer system, the fourth natural language utterance using the corresponding one or more contexts in the updated context stack; and 
 generating, by the computer system, a response to the fourth natural language utterance based on the interpretation of the fourth natural language utterance. 
 
     
     
       6. The method of  claim 1 , the method further comprising:
 accumulating, by the computer system, short-term knowledge, wherein the short-term knowledge is accumulated based on one or more of the series of natural language utterances received by the computer system during a predetermined time period. 
 
     
     
       7. The method of  claim 6 , wherein the determination that the third natural language utterance corresponds to one or more of the series of contexts in the generated context stack is based on the short-term knowledge. 
     
     
       8. The method of  claim 6 , the method further comprising:
 responsive to a determination that the third natural language utterance does not correspond to one or more of the series of contexts in the generated context stack, determining, by the computer system, a third context for the third natural language utterance based on the short-term knowledge; 
 interpreting, by the computer system, based on the third context, the third natural language utterance; and 
 generating, by the computer system, a response to the third natural language utterance based on the interpretation of the third natural language utterance based on the third context. 
 
     
     
       9. The method of  claim 1 , wherein: the series of contexts further comprises a fourth context identified for a fourth utterance at a fourth time after the second time and before the third time; and, in the generated context stack, the fourth context is listed before the second context. 
     
     
       10. A system for facilitating natural language system responses utilizing a context stack generated based on multiple prior natural language utterances received, the system comprising:
 one or more physical processors programmed with one or more computer program instructions which, when executed, configure the one or more physical processors to:
 track a series of contexts respectively identified for a series of natural language utterances received by the system during a current conversation, the series of contexts comprising at least a first context identified for a first utterance at a first time and a second context identified for a second utterance at a second time after the first time; 
 generate a context stack based on the tracked contexts comprising the series of contexts in reverse chronological order of a time of identification by the computer system, wherein the second context is listed before the first context; 
 receive a third natural language utterance from an input device as part of the current conversation at a third time after the second time; 
 determine whether the third natural language utterance corresponds to one or more of the series of contexts in the generated context stack by comparing the third natural language utterance to one or more of the series of contexts in the order the series of contexts are listed in the generated context stack; and 
 responsive to a determination that the third natural language utterance corresponds to one or more of the series of contexts in the generated context stack, interpret the third natural language utterance using the corresponding one or more contexts. 
 
 
     
     
       11. The system of  claim 10 , wherein the one or more physical processors are further configured to:
 generate a response to the third natural language utterance based on the interpretation. 
 
     
     
       12. The system of  claim 10 , wherein to determine whether the third natural language utterance corresponds to one or more of the series of contexts in the generated context stack, the one or more physical processors are configured to:
 identify one or more of the series of contexts in the generated context stack based on one or more recognized words of the third natural language utterance. 
 
     
     
       13. The system of  claim 10 , wherein to determine whether the third natural language utterance corresponds to one or more of the series of contexts in the generated context stack, the one or more physical processors are configured to:
 determine whether the third natural language utterance corresponds to a most recent context of the series of contexts in the generated context stack; and 
 responsive to a determination that the third natural language utterance does not correspond to the most recent context, determine whether the third natural language utterance corresponds to a second most recent context of the series of contexts in the generated context stack. 
 
     
     
       14. The system of  claim 10 , wherein the one or more physical processors are further configured to:
 responsive to a determination that the third natural language utterance does not correspond to one or more of the series of contexts in the generated context stack, determine a third context for the third natural language utterance based on one or more recognized words of the third natural language utterance; 
 update the generated context stack based on the third context by placing the third context first; 
 receive a fourth natural language utterance; 
 responsive to a determination that the fourth natural language utterance corresponds to one or more of the series of contexts in the updated context stack, interpret the fourth natural language utterance using the corresponding one or more contexts in the updated context stack; and 
 generate a response to the fourth natural language utterance based on the interpretation of the fourth natural language utterance. 
 
     
     
       15. The system of  claim 10 , wherein the one or more physical processors are further configured to:
 accumulate short-term knowledge, wherein the short-term knowledge is accumulated based on one or more of the series of natural language utterances received by the system during a predetermined time period. 
 
     
     
       16. The system of  claim 15 , wherein the determination that the third natural language utterance corresponds to one or more of the series of contexts in the generated context stack is based on the short-term knowledge. 
     
     
       17. The system of  claim 15 , wherein the one or more physical processors are further configured to:
 responsive to a determination that the third natural language utterance does not correspond to one or more of the series of contexts in the generated context stack, determine a third context for the third natural language utterance based on the short-term knowledge; 
 interpret the third natural language utterance based on the third context; and 
 generate a response to the third natural language utterance based on the interpretation of the third natural language utterance based on the third context. 
 
     
     
       18. The system of  claim 10 , wherein: the series of contexts further comprises a fourth context identified for a fourth utterance at a fourth time after the second time and before the third time; and, in the generated context stack, the fourth context is listed before the second context.